Towards more Practical Threat Models in Artificial Intelligence Security

Chat with Graph Search

Ask any question about EPFL courses, lectures, exercises, research, news, etc. or try the example questions below.

DISCLAIMER: The Graph Chatbot is not programmed to provide explicit or categorical answers to your questions. Rather, it transforms your questions into API requests that are distributed across the various IT services officially administered by EPFL. Its purpose is solely to collect and recommend relevant references to content that you can explore to help you answer your questions.

Developing agents that can reliably act on our behalf is central to artificial intelligence (AI). These agents must seamlessly interact with tools, like search engines and databases, and collaborate. In this thesis, we study the abstractions, methods, and ...

Early work has found that large language models (LLMs) can generate persuasive content. However, evidence on whether they can also personalize arguments to individual attributes remains limited, despite being crucial for assessing misuse. This preregistere ...

Driving continuous, low-power artificial intelligence (AI) in the Internet of Things (IoT) requires reliable energy harvesting and storage under indoor or low-light conditions, where batteries face constraints such as finite lifetimes and increased environ ...

Faithful human performance capture and free-view rendering from sparse RGB observations is a long-standing problem in Vision and Graphics. The main challenges are the lack of observations and the inherent ambiguities of the setting, e.g. occlusions and dep ...

In contrast to vast academic efforts to study AI security, few real-world reports of AI security incidents exist. Released incidents prevent a thorough investigation of the attackers' motives, as crucial information about the company and AI application is ...

Video DeepFakes are fake media created with Deep Learning (DL) that manipulate a person’s expression or identity. Most current DeepFake detection methods analyze each frame independently, ignoring inconsistencies and unnatural movements between frames. Som ...

MammAlps is a multimodal and multi-view dataset of wildlife behavior monitoring. Nine camera-traps were placed at three monitoring sites in the Swiss National Park, from which we curated over 14 hours of video with audio, 2D segmentation maps and 8.5 hours ...

This article reports on the current state of the OBI DICT project, a bilingual e-dictionary of oracle-bone inscriptions (OBI), incorporating artificial intelligence (AI) image recognition technology. It first provides a brief overview of the development of ...

To retrieve Surface Solar Radiation (SSR) from satellite images, a baseline reflectance of the observed ground in clear-sky conditions, also called background reflectance, is necessary to distinguish atmospheric absorption and scattering effects from surfa ...

Assessing the individual risk of Major Adverse Cardiac Events (MACE) is of major importance as cardiovascular diseases remain the leading cause of death worldwide. Quantitative Myocardial Perfusion Imaging (MPI) parameters such as stress Myocardial Blood F ...

The optimistic gradient method is useful in addressing minimax optimization problems. Motivated by the observation that the conventional stochastic version suffers from the need for a large batch size on the order of O(ϵ-2) to achieve an ϵ-stationary solut ...

Driven by the demand for real-time processing and the need to minimize latency in AI algorithms, edge computing has experienced remarkable progress. Decision-making AI applications stand out for their heavy reliance on data-centric operations, predominantl ...

Monitoring of cardiac output (CO) is a mainstay of hemodynamic management in the acutely or critically ill patient. Invasive determination of CO using thermodilution, albeit regarded as the gold standard, is cumbersome and bears risks inherent to catheteri ...

Deep learning models for learning analytics have become increasingly popular over the last few years; however, these approaches are still not widely adopted in real-world settings, likely due to a lack of trust and transparency. In this paper, we tackle th ...

Adversarial examples crafted in black-box scenarios are affected by unrealistic colors or spatial artifacts. To prevent these shortcomings, we propose a novel strategy that generates adversarial images with low detectability and high transferability. The p ...

According to the proposed Artificial Intelligence Act by the European Comission (expected to pass at the end of 2023), the class of High-Risk AI Systems (Title III) comprises several important applications of Deep Learning like autonomous driving vehicles ...

The convergence of human and artificial intelligence is currently receiving considerable scholarly attention. Much debate about the resulting Hybrid Minds focuses on the integration of artificial intelligence into the human brain through intelligent brain- ...

Deepfakes - synthetic videos generated by machine learning models - are becoming increasingly sophisticated. While they have several positive use cases, their potential for harm is also high. Deepfake production involves input from multiple engineers, maki ...

Explainable Artificial Intelligence (XAI) plays a crucial role in enabling human understanding and trust in deep learning systems, often defined as determining which features are most important to a model's prediction. As models get larger, more ubiquitous ...

Harnessing the power of machine learning (ML) and other Artificial Intelligence (AI) techniques promises substantial improvements across forensic psychiatry, supposedly offering more objective evaluations and predictions. However, AI-based predictions abou ...

Towards more Practical Threat Models in Artificial Intelligence Security

Chat with Graph Search

Collaborative AI Agents in the Era of Large Language Models

On the conversational persuasiveness of GPT-4

Unlocking high-performance photocapacitors for edge computing in low-light environments

MetaCap: Meta-learning Priors from Multi-view Imagery for Sparse-View Human Performance Capture and Rendering

When Your AI Becomes a Target: AI Security Incidents and Best Practices

Efficient Temporally-Aware DeepFake Detection using H.264 Motion Vectors

MammAlps: A multi-view video behavior monitoring dataset of wild mammals in the Swiss Alps

The Current State of the OBI DICT Project: A Bilingual e-Dictionary of Oracle-Bone Inscriptions with AI Image Recognition

Retrieval of Surface Solar Radiation through Implicit Albedo Recovery from Temporal Context

Comparing various AI approaches to traditional quantitative assessment of the myocardial perfusion in [82Rb] PET for MACE prediction

Diffusion Stochastic Optimization for Min-Max Problems

EdgeAI-Aware Design of In-Memory Computing Architectures

Machine learning-enabled estimation of cardiac output from peripheral waveforms is independent of blood pressure measurement location in an in silico population

Trusting the Explainers: Teacher Validation of Explainable Artificial Intelligence for Course Design

GLoFool: Global Enhancements and Local Perturbations to Craft Adversarial Images

Robust Training and Verification of Deep Neural Networks

Playing Brains: The Ethical Challenges Posed by Silicon Sentience and Hybrid Intelligence in DishBrain

A multimodal measurement of the impact of deepfakes on the ethical reasoning and affective reactions of students

The future of human-centric eXplainable Artificial Intelligence (XAI) is not post-hoc explanations

Out of their minds? Externalist challenges for using AI in forensic psychiatry